Expressive writing about stressful experiences has beneficial effects on some psychological, physiological, and symptom outcome measures. Recently, this intervention has been demonstrated to produce disease improvement in patients with organic disease. We propose to examine expressive writing as complementary treatment for individuals with diabetes mellitus. An estimated 16 million people have diabetes, with a 33% increase in the prevalence of diabetes between 1990 and 1998. Diabetes remains a leading cause in the US of blindness, end-stage renal failure and nontraumatic lower extremity amputation. This study addresses the following questions: 1) Does the benefit of expressive writing extend to patients with diabetes? 2) What are the potential mediators and moderators of any observed effects? 3) What is the duration of any observed improvements? 4) Does the application of additional expressive writing "booster" sessions lead to greater improvement and/or a more sustained improvement? The study is a multi-site, prospective, randomized experimental design. Patients (n=225) with diabetes will be recruited from three geographically distinct areas, and randomly assigned to the control group (neutral writing) or one of two experimental conditions (expressive writing or expressive writing plus booster). Participants in the experimental conditions complete a written expression exercise, writing for several sessions about traumatic or stressful events, while control group participants write about neutral topics. One experimental group will receive "booster" sessions of additional writing at 4-month intervals. All participants will have disease status, quality of life, and psychological factors monitored every 4 months for a period of two years